(oe) 
WM. H. HOBBS 
rocks, descriptions of which have been placed onrecord. The lines 
separating the several types are fixed arbitrarily, and would, in 
general, be located somewhat differently if undertaken at the 
outset by different individuals. For the types of larger order, 
these lines have been fixed by the traditional rock groups, and 
they are not likely to be much changed, but for the new and 
specific types they will be largely determined by the particular 
rock areas which are first examined. 
A much more general use of intermediate family type names 
is inevitable, and terms like grano-diorite (or better, granito- 
diorite), trachy-andesite, etc., should be utilized." 
Rock relationships should be indicated by the combination of names 
into a binominal, or, if necessary, polynominal nomenclature.—The 
multiplication of specific terms, whose derivation has only a 
geographical signification (¢. g., Toscanite, Absarokite, Litch- 
fieldite), furnishing not the slightest indication of the rock’s 
relationships, is fast bringing petrologists to the condition of 
the Chinaman who is required to learn a unique syllable for 
every word in his language. Not possessing the admirable 
memory training of the Chinaman, the petrographer finds him- 
self somewhat bewildered under the rain of new petrographical 
names which has characterized the closing years of the century. 
Many of these terms have been rendered necessary by the elabo- 
ration of the system of classification, due to the improved 
methods of chemical examination, and to the discovery of new 
petrographical provinces, and others are sure to be needed, but 
the enterprise in this branch of the science manifested in some 
quarters has sometimes provided us with two, or even three, 
names for the same specific rock type. 
There can be no question that the nomenclature of petrog- 
raphy can be greatly simplified by a return to a binomial or 
polynomial nomenclature, which, fortunately, can be accom- 
plished without much confusion, provided the old names of rock 
families be retained, together with compound names for the 
gradational types connecting them. An illustration may be 
Cf. BROGGER: op. cit., p. 93. 
